The current proposal is for a follow-up of subjects originally studied in a cohort-sequential investigation of adolescent cigarette smoking. Cohorts of 6th-12th graders (total N=8681) were originally followed between 1980-1983 with the goal of prospectively predicting adolescent smoking initiation from several categories of social psychological factors. At follow-up, these subjects will range from 10th graders to young adults (mid 20s). The proposed follow-up extends our original work by linking adolescent and adult cigarette smoking. Continuities and discontinuities between adolescent and adult tobacco use will be determined. In addition, we will examine the utility of our social psychological factors in predicting adult smoking for adolescents who are and are not regular smokers. Moreover, using our cohort-sequential design, we will track changes in smokling attitudes, beliefs, and perceived environments that occur with increasing age and with cultural-historical change. We will determine whether the high school milieu has become more "antismoking" over the seven year period (1980-1987). Finally, we will answer a methodological question by determining the accuracy of retrospective reports of smoking behavior. Taken together, the original data base along with the proposed follow-up form a comprehensive picture of the natural history of smoking from the beginning of the period of risk (pre-adolescence) to the end of the period of risk (young adulthood). The data have important implications for the design of antismoking campaigns by identifying high risk target audiences and the factors that are important to long term smoking outcomes.